INTRODUCTION

TO ANALYTICS
2022 - 2023
 LESSON 3.
ANALYTICS PRIMER:
GOALS, APPLICATIONS & TYPES OF
ANALYTICS
 Learning Objectives

1. Articulate goals of analytics

2. Recognize how analytics is used in real world
3. Justify business benefits of analytics
4. Identify analytics application in different domains
5. Distinguish between descriptive, diagnostic, predictive and
prescriptive analytics
6. Practice asking analytics questions
Agenda

1. Goals, applications & benefits of analytics

2. Analytics domains; future trends
3. Types of analytics
4. Analytics questions
 ANALYTICS GOALS,
APPLICATIONS, DOMAINS
What do we do with data?

Discuss the article given out as home assignment.

What data was collected?

What was done with the collected data?

How was it used?

 Goal of analytics

Analytics: The examination of information to uncover insights that give a

businessperson the knowledge to make informed decisions.

https://www.ibm.com/developerworks/community/blogs/jfp/entry/the_analytics_maturity_model?lang=en
Benefits of Analytics

Textbook
Chapter 2
Figure 2.1
Using analytics
Using analytics
Using analytics
Using analytics
Analytics Applications Examples

• Recommender systems - based on previous search results for a user

• Targeted advertising - targeted based on user’s past behavior; social media
analytics
• Speech recognition – sound patterns are compared to find best match; predict
intention of speech
• Typing autocomplete – predicting the words based on the context and first letters
• Gaming - algorithms which improve / upgrade themselves as the player moves up
to a higher level
• Fraud and risk detection – likelihood to default on loan is calculated based on
customer profile
• Logistics - best routes to ship, best time to deliver, best mode of transport
• Sports cameras – real-time prediction of game direction and switching cameras
 Analytics Domains
Marketing: Security:
• Targeted marketing • Spam filtering
• Online advertising • Suspicious behaviour detection
• Cross-selling recommendations • Data access pattern monitoring

Customer relationship: Human resources:

• Reduce attrition • Performance measurement
• Predict customer need: next best action • Employee satisfaction
• Detect or predict problems • Hiring and retention

Retail:
Finance:
• Supply chain management
• Credit scoring
• Demand prediction
• Trading
• Market basket analysis
• Fraud detection
Analytics Trends

• Self-driving cars
• Sentiment analysis
• Face recognition
• Natural language processing
• Robots
• Endless….
TYPES OF ANALYTICS
 Types of Analytics

Descriptive The examination of data or content, to answer the question

Analytics “What happened?” or “What is happening?”

Diagnostic A form of advanced analytics that examines data or content to

analytics answer the question, “Why did it happen?”

Predictive A form of advanced analytics which examines data or content

Analytics to answer the question “What is going to happen?” or more
precisely, “What is likely to happen?”

Prescriptive A form of advanced analytics which examines data or content

Analytics to answer the question “What should be done?” or “What can we
do to make X happen?”
https://www.gartner.com/en/information-technology/glossary
Analytics Value and Focus
Traditional (Descriptive) Analytics

2. Infer why has it

occurred

1. Monitor what has

occurred
Descriptive Analytics

Questions “What happened?”, “What is happening?”

Goals Make data from multiple data sources visible to human decision makers;
describe data samples through summarization and visualization.
Describes the world as it is. Provides insights into the past and present

Tools Business intelligence (BI) and visualization tools:

Spreadsheets, pie charts, bar charts, line graphs, tables, OLAP,
dashboards and reports, data exploration and visualization

Examples Google Analytics

Diagnostic Analytics

Questions “Why did it happen?”, “What are the trends?”, ”What patterns are
there?”

Goals Measure historical data, compare it to other data to find patterns,

dependencies and identify trends.
Provides insights into the past.

Tools Drill-down, data discovery, data mining and correlations

Examples Market segments, sentiment analysis, root cause for failures, IT log
analytics, price elasticity, fraud patterns, scientific research
Advanced Analytics 2. Determine the
best course of
action

1. Predict and
forecast future
Predictive Analytics

Questions “What is likely to happen?”

Goals Examine findings of descriptive and diagnostic analytics to predict new data
about present and future.

Tools Regression analysis, forecasting, multivariate statistics.

Unique algorithms are created from data sets to build predictive models

Examples Predict customer behavior (buy propensity), customer churn propensity

(likelihood that a customer leaves for a competitor), fraud score (likelihood of a
credit card transaction to be fraudulent), extract meaning from free text (what
is the most likely meaning)
Prescriptive Analytics

Questions “What should we done?” or “What can we do to make it happen?”

Goals Prescribe what action to take to eliminate a future problem or take advantage
of an opportunity.
Determine what actions should be executed now (operational decisions) or in
the future (tactical and strategic decisions).

Tools Graph analysis, optimization, simulation, complex event processing, neural

networks, recommendation engines, heuristics.
Leverages big data, algorithms ands business rules.
Relies on machine learning and deep learning.

Examples Demand forecast, inventory management

Cognitive Analytics – the next level

Cognitive The future state of AI

analytics Human-like decision-making
Advanced machine intelligence with self-learning capabilities

Self- Draw inferences from existing data and patterns

learning Draw conclusions based on existing knowledge
Add new conclusions (new knowledge) back into the knowledge base for
future inferences
 Later in the program

Quantitative Research Using quantitative methods to generate descriptive statistics

Methods Course

Predictive Analytics Methods applied to create predictive models

Course
Types of Analytics:

let’s check our understanding

A. Descriptive
B. Diagnostic
C. Predictive
D. Prescriptive
ANALYTICS
QUESTIONS
The importance of asking the right questions
The importance of asking the right questions

https://www.linkedin.com/feed/update/urn:li:activity:6887906139858591744/
How to ask analytics questions?

Work in groups:

A streaming platform has collected a lot of data

about user behaviour on the platform, as well as
customer surveys.

What analytics questions could we ask to improve

the platform movie recommendations algorithm?
 Descriptive Analytics

Cumulative (counter, Number of products sold last month, monthly data usage
volume)
Total followers, Total rainfall

Delta (change) Sales growth

Defect reduction

Gauge (value measured CPU usage

at a moment in time)
Speed
Sound volume
Rate of occurrence Infections per million population
Accidents per million miles driven

Descriptive statistics Central tendency (mean, median, mode),

Measures of variation and distribution
 Descriptive Analytics

Aggregate Percentage of customers by age group, income level or gender

analysis
Total products sold in each product category
Profitability of each product category
Trend analysis Sales trends
Customer growth trend
Stock price trend
Contribution Contribution to sales volume by product category
analysis
Movies watched by customer by genre
Key Customer retention rate
performance
indicators Net profit
Employee satisfaction score
25 Need-To-Know Key Performance Indicators (Bernard Marr, 2014)
FT Press
 Diagnostic Analytics

Correlation Correlation between two observations (positive or negative correlation)

analysis

Hypothesis Prove or disprove an assumption e.g. identify likely reasons behind customer
testing churn increase, product popularity, or customer dissatisfaction

Regression Relationship between two or more variables e.g. number of accidents vs. rainfall
analysis vs. time of the day

Cohort analysis Comparison of metrics of different cohorts (groups) e.g. spending patterns,
reading preferences by demographic group
 Good Analytics Questions

Clear Was there any difference in how many sales

we’ve made last year considering that we
have added two new products and including
each country that we exited, except for our
leading products?
Specific What did we do well last year?

Use accurate terms How many widgets were produced lately?

Feasible to get How many people will live on Mars by 2300?

answered

The answer can be What party is most likely to be in power in

used 2068?